Quantitative Descriptive Study Design Descriptive Correlational Research By

Quantitative Descriptive Study Design: Descriptive Correlational Research

By the completion of this presentation, the participant will be able to: • Describe three characteristics of a descriptive study • Explain two components of a correlational study • Discuss the major strengths and weaknesses for one type of descriptive study

Research Design Blueprint • Action plan for conducting research study – Techniques and procedures – Reduces researcher bias – Controls for extraneous variables – Controls for other sources of variances • Produces credible, high quality research findings – Clear and detailed: • Understand study aim & purpose • How research was conducted • Evaluate the research process • Reproduce research study

Robust Research Design • Considers the following: –Environment – Equivalence – Treatment – Measurement – Extraneous variables – Data analysis

Research Design APPROACH TYPES DESIGN QUALITATIVE (discovers) PHENOMENOLOGICAL GROUNDED THEORY ETHNOGRAPHIC HISTORICAL PHILOSOPHICAL QUALITATIVE or QUANTITATIVE (describes) CORRELATIONAL DESCRIPTIVE Non-experimental (Observational) QUANTITATIVE (explains; cause & effect) EXPERIMENTAL QUASIEXPERIMENTAL Experimental

QUANTITATIVE RESEARCH CHARACTERISTICS PHILOSOPHY TECHNIQUES Hard Science New Knowledge Concise and limited focus Tests Theory Reductionistic Explains & Predicts Objective Uses Instruments Logistic and Deductive Reasoning Numbers Statistical Analysis Generalization

TERMINOLOGY RESEARCH APPROACH SOCIAL SCIENCE TERM EXPERIMENTAL MEDICAL RESEARCH TERM Randomized Control Trial; Randomized Clinical Trial (RCT) QUASIEXPERIMENTAL Controlled Trial; Controlled Trial without randomization NONEXPERIMENTAL: DESCRIPTIVE CORRELATIONAL RETROSPECTIVE PROSPECTIVE Observational Studies Case-Series Case-Control Cross-Sectional Cohort Prevalence Incidence QUANTITATIVE

Non-experimental Quantitative Research Descriptive Design • Describes phenomena in real life • Describes, groups, and classifies situations that does not concepts manipulate variables • Examines characteristics of a single sample in order to generalize to a single population • Generates new knowledge when little or no knowledge is available • Develops models & theories

Descriptive Research Design Case Study • In-depth analysis and systematic description of one patient or one group of similar patients • No manipulation of variables • Common in nursing 40 -50 years ago, but are now less frequent • Can be used: – as evidence to support or invalidate theories – to generate new hypotheses for testing – to demonstrate effectiveness of Therapeutic techniques

Descriptive Research Design Case Study: Advantages • Wealth of detail • Wide variety of information • Clues & ideas for further research • Understand a topic, concept, issue in general in order to study it in greater detail Case Study: Advantages Hard to – control – do well and easy to do badly – tell if it has been done badly One is often unable to determine if researcher examined – the most important topic, clues – what was eliminated or not Conclusions only apply the one case – Does not create conclusions beyond the one case. Disadvantages explains why case studies are now out of vogue Subjective: Researcher decides – What to look for or ignore – What to record or not record – What is important or not – What clues to follow or drop

Descriptive Research Design Case-Series Design • Simple descriptive account of interesting characteristics seen in a group of people –Short time period –Important descriptive role as a precursor to designing other research studies to: • Evaluate causes • Explanations of observations –Acknowledges observed characteristic bias

Descriptive Research Design Cross-Sectional Design • Also called Survey, Prevalence, Incidence, Epidemiologic Studies • Diagnose or stage a disease • Analyze data collected on a group at one time period – Subject and information obtained in a short time frame • Usefulness of new or current diagnostic procedures • What is happening • Establishing norms • Gain insight into a topic or learn people’s perceptions (surveys) Most common research design in nursing

Non-experimental Quantitative Research Descriptive Correlational Design • First: Describes variables • Second: Examines relationships amongst these variables – Does not infer cause-and-effect relationships • Facilitates the identification of many interrelationships in a particular situation – Situation may have occurred or is currently occurring – No attempt to control or manipulate the situation

Non-experimental Quantitative Research Correlational Design • Primary purpose is the examination of relationships – Examine relationships between 2 or more variables • No manipulation of variables Determine if a relationship exists between variables – None, weak, moderate, or strong • Determine type of relationship between variables – Positive relationship or negative relationship

Correlational Research Design Case-Control Design • Retrospective in nature –Presents phenomenon linked to past phenomenon: “What has happened? ” –Looking back in time to detect causes or risk factors for the presence or absence of an outcome • Example: Case-Control Studies –Cigarette smoking lung cancer –People with lung cancer = cases –People without lung cancer = controls • Differences between groups = smoking

Correlational Research Design One Group Designs: Single Group • Descriptive design : no random selection of subjects – Convenience or volunteer sample • Examine characteristics of a single group • Natural setting • Measurements made about the group Subjects serve as own control – Measure group X 1, X 2, or more – Still one group EXAMPLE: – Satisfaction surveys one group over 12 month time period - Use the resulting descriptions to draw conclusions about that groups’ satisfaction scores

Correlational Research Design One Group Designs: Time Dimension Interrupted Time Series • Involves more than one pre and post test measurement – Equal number of measurements before and after the intervention – The time periods must be constant and equal Longitudinal Time Series Follows one group to examine and measure changes in same subjects over an extended time period

Correlational Research Design One Group Designs: Time Dimension Strengths • Allows examination of sequences and patterns of change over both – single time period – Interrupted time periods Weaknesses • Multiple measurements over an extended period • Subject drop out – Threatens instruments validity & reliability • Distorted data – Subject keeps a copy of the survey and duplicated responses • Hawthorne effect

Correlational Research Design One Group Design: Summary • Correlational • No random assignment • Characteristics of single sample • Pre test/post test • Time series or multi-variant • Longitudinal

Non-experimental Quantitative Research Comparative Descriptive Design • Describes variables & examines differences in 2 or more groups – Occurs naturally in a setting – No manipulation of variables • Results obtained from the final analysis are frequently not generalized to a population

Correlational Research Design Multiple Group Designs Group comparative designs – Simple 2 group design – Post-test design – Pre/Post Test Design – Time Series Design ADVANTAGES: • Comparison of group on dependent variables • Examines differences between groups • Examines differences within groups – Are they coherent groups? – Are there true differences between groups? – Did change occur re: multiple variables within groups?

Correlational Research Design Multiple Group Designs: Multiple Group Time Series Memory loss in the Elderly: • Multiple independent variables • One dependent variable – measured as multiple pre & post tests • Constant & equal time periods • Groups: (1) Ginkgo Bilbo (2) Plavix, & (3) No medication (control) • Measure via memory test (Pre-test) • Low dose & high dose of medications • Measure via memory test (Post-test) • Repeat • Which group demonstrates the best memory scores

Correlational Research Design Multiple Group Designs: Multiple Group Time Series • Compare 2 or > groups on natural phenomenon • Sometimes called descriptive studies • Sometimes called inferential studies • Concerns are raised when comparing naturally occurring phenomenon in multiple groups – May be better addressed by using quasi-experimental procedures that are inferential, as opposed to correlational procedures

Summary & Conclusions Descriptive Correlational Design • Philosophical commonalities Research techniques • Lack of variable manipulation & variable control – Can result in possible bias – see Module 10! • Partner with seasoned researchers and statisticians – Develop competencies needed to conduct quality research studies – Produce credible research findings – Impact patient care